Berberine and “nature’s Ozempic”: what the weight trials and safety reviews show

Berberine is a yellow plant alkaloid found in barberry, goldenseal and Oregon grape. It has been sold for years for blood sugar and cholesterol. Since 2023, social media has sold it as “nature’s Ozempic”, a supplement stand-in for GLP-1 weight-loss medicines. The weight-loss trials show a small average effect. In a European safety review, EFSA’s panel could not set a safe intake for the plant preparations it assessed.

Neither finding is as dramatic as the marketing, and the sources cited here do not report a ban. EFSA’s text is scientific advice out for public comment. It is not a final opinion, and it is not a rule in the UK, US, UAE or South Africa.

The short answer: berberine is not a GLP-1 medicine, and its average weight effect is small. A 2026 meta-analysis of 23 randomised trials found 0.88 kg more weight loss than control. STEP 1, a different and longer semaglutide trial, found a much larger difference versus placebo. EFSA’s panel could not set a safe intake for the 13 plant preparations it assessed. Goldenseal and greater celandine have reported harms. For most of the other plants the gap is missing safety studies, not a demonstrated harm.

What “nature’s Ozempic” gets wrong

Semaglutide, the ingredient in Ozempic and Wegovy, acts on the GLP-1 receptor. In a news release of 11 March 2026, Wroclaw Medical University said berberine does not act as a hormonal drug and does not activate one specific receptor, and that its metabolic effects are indirect and depend on the gut microbiota. That comparison is the university’s commentary on its gut review, not a line from the paper itself. NutraIngredients reports that berberine went viral on TikTok in 2023 under #naturesozempic.

What the weight-loss trials show

• A 2026 systematic review and meta-analysis in the International Journal of Obesity pooled 23 randomised trials in adults. Berberine lowered body weight by 0.88 kg more than control (95% confidence interval -1.36 to -0.39) and BMI by 0.48 kg/m² (-0.89 to -0.07). The waist-to-hip ratio did not change significantly. The authors call for better reporting of purity, potency and gram amounts, and note common biases such as lack of blinding and randomisation.
• An earlier meta-analysis of 12 trials (Clinical Nutrition ESPEN, 2020) found a larger average weight difference of 2.07 kg, with BMI 0.47 kg/m² lower. The two reviews pooled different trials and disagree on the size of the effect.
• The US National Center for Complementary and Integrative Health says the evidence is not conclusive. Citing a 2022 review, it says weight effects were seen mainly above 1 g a day for more than 8 weeks, that many included studies had a high risk of bias, and that most participants had health problems such as diabetes or fatty liver and were studied in Asia, with very few in North America.
• A 12-week Greek trial in Nutrients (August 2026) gave 93 adults with obesity either a combination of berberine (1,000 mg), white mulberry leaf extract and chromium picolinate, or placebo. The combination group had a greater fall in BMI (-1.24 kg/m², -2.15 to -0.33) and total fat mass (-2.35 kg, -4.12 to -0.58), and a 28% greater improvement in insulin sensitivity (ratio 1.28). There was no berberine-only arm. The effect cannot be attributed to berberine alone.

For scale only, the STEP 1 trial of once-weekly semaglutide 2.4 mg in 1,961 adults with overweight or obesity found a mean weight change of -14.9% over 68 weeks against -2.4% on placebo, a reported between-group difference of 12.4 percentage points, and a mean loss of 15.3 kg against 2.6 kg, a difference of 12.7 kg. Those rounded percentages do not by themselves produce the 12.4-point figure. The kilogram difference is 15.3 minus 2.6. STEP 1 is not a head-to-head trial with berberine. The populations and the 68-week duration differ. On current evidence, berberine’s average effect is small, and it cannot honestly be described as a substitute for a GLP-1 medicine.

What European safety bodies found

France’s ANSES published an opinion on 1 August 2019, after a request from the DGCCRF, the French directorate for competition, consumer affairs and fraud control. It said berberine’s pharmacological actions are confirmed at 400 mg a day or more and cannot be excluded at lower doses. With poor-quality toxicology studies, it set only an indicative toxicity value of 1.7 micrograms per kg of body weight per day, 0.1 mg a day for a 60 kg adult. Those figures are not a dose to take. It said pregnant and breastfeeding women, people with diabetes, people with liver or heart disorders, and children and adolescents should avoid berberine supplements, and it warned of drug interactions, notably with cancer treatments. It found no scientific basis for Belgium’s 10 mg a day of isoquinoline alkaloids, expressed as berberine, a line those authorities used to separate supplements from medicines, not a universal safe dose.

At the European Commission’s request, EFSA’s nutrition panel assessed 13 berberine-containing plants used in supplements, including Berberis aristata, Berberis vulgaris, goldenseal and greater celandine. It endorsed the text for public consultation on 29 January 2026. Slides for food businesses on 22 April 2026 still said the consultation ran to 11 May 2026, with possible adoption on 1 or 2 July. SupplySide Supplement Journal reported on 3 July 2026 a later extension, at the Commission’s request, to 10 July 2026. That date is the magazine’s report, not an adopted rule. Benefits, medicines and synthetic berberine were out of scope. The plants were assessed as preparations, which contain other constituents as well as berberine. The slides say:

• There is evidence of berberine genotoxicity in vitro, meaning gene mutation and chromosomal damage in laboratory tests, supported by mechanistic evidence. Confirmation in living organisms is lacking, especially in the gut and liver.
• There is no adequate repeated-dose toxicity study on berberine alone from which a reference point could be identified.
• Berberine supplements may cause transient gut symptoms: constipation, diarrhoea, nausea and abdominal pain.
• Berberine may inhibit CYP3A4 and possibly other CYP450 enzymes that process medicines, indicating a risk of interaction with various medicines.
• Preparations from goldenseal rhizome and root showed carcinogenic activity in rodents, particularly liver adenomas, and EFSA concluded that they represent a carcinogenic risk for humans. Greater celandine herb has been linked to idiosyncratic liver injury in people. Susceptible individuals cannot currently be identified, and no dose below which such reactions would not occur can be set. The slides say the role of berberine in those liver injuries remains uncertain.
• Except for goldenseal, the toxicity profile of the other plant preparations beyond their berberine content is largely unknown because adequate toxicity studies are missing. Overall, the data do not allow a safe intake to be established for any of the preparations assessed.

EFSA’s exposure figures are theoretical berberine intakes, not measured diets. From a call for data, 47 analytical samples of Berberis aristata bark or root gave a median 445 mg a day and a maximum 496 mg, using measured content times the recommended daily amount. In the Mintel database, 30 products gave a median 274 mg a day and a maximum 1,000 mg, assuming the label content is accurate. Both assume people follow the recommended amount. There is no actual consumption survey behind them.

How strong is the safety case?

The UK Committee on Toxicity discussed, on 31 March 2026, the berberine text EFSA had put out for public comment. The minutes, published on 30 June 2026, record concern that the evidence indicated a possible genotoxic hazard and was not strong enough to rule that hazard out. They also record that EFSA’s QSAR analysis used only VEGA tools and was not robust enough to support a genotoxicity conclusion, that the exposure estimates appeared largely theoretical, and that members doubted the rat was a good model for human metabolism. The committee concluded that substantial gaps on genotoxicity, exposure estimation and toxicokinetics reduced confidence in those conclusions. That is not a statement that berberine is safe.

Ingredient suppliers told NutraIngredients they want any restriction matched to each preparation, and that a missing study is not the same thing as a demonstrated hazard. That describes the data gap for preparations other than goldenseal. It does not make berberine proven safe. There is still no adequate long-term toxicity study from which to judge use around 1 g a day, and both EFSA and the National Center for Complementary and Integrative Health flag interactions with medicines.

EFSA’s text is scientific advice still in public consultation, not a final opinion and not a ban. Decisions on any restriction belong to risk managers, in the Commission and member states. NutraIngredients reported in July 2026 that a regulatory expert expected EFSA’s final opinion in early 2027, with a Commission decision to follow. That expectation is the magazine’s, not an adopted rule. EFSA’s consultation text and ANSES’s advice are not rules in the UK, US, UAE or South Africa.

Medicines, including GLP-1 medicines

ANSES names metformin, other diabetes medicines, medicines with a narrow safety margin, and cancer treatments among the interactions. No source cited here reports a trial of berberine with a GLP-1 medicine. The National Center for Complementary and Integrative Health says berberine interacts with cyclosporine and that people who take medicines should talk to a doctor or pharmacist. It also says berberine is likely unsafe for infants, and may be unsafe in pregnancy or while breastfeeding, because of a harmful build-up of bilirubin. Ask a pharmacist before adding it to any of these medicines.

If you are still considering it

• The pooled trials show a small average weight change. They are not a reason to start berberine, and they are not a GLP-1 effect.
• A label that says only “berberine” may hide the plant. Goldenseal and greater celandine carry the signals above. Synthetic berberine was outside the plant-preparation assessment.
• ANSES says to avoid berberine supplements in pregnancy, while breastfeeding, in children and adolescents, and with diabetes or a liver or heart disorder. If you take regular medicines, speak to a pharmacist or doctor first. Do not combine several supplements. The 400 mg, 0.1 mg and Belgian 10 mg figures are not a personal dose.

UAE rules

EFSA and ANSES do not set rules in the UAE. In Dubai, no health supplement may be manufactured, imported, exported, advertised, sold or distributed unless it is registered in Dubai Municipality’s Montaji system, and shoppers can check a product by scanning the barcode in the Montaji app. That is a Dubai record, not approval of every product in the country, and not evidence of an effect on weight or blood sugar. Federally, MOHAP has transferred a defined set of services to the Emirates Drug Establishment, including manufacturer registration and renewal of some pharmaceutical products. The UAE government portal says the Establishment is to regulate medical products, including food supplements. A Montaji record is not that federal mandate.

FAQ

Is berberine the same as Ozempic?

No. Ozempic contains semaglutide, which activates the GLP-1 receptor. Berberine is a plant alkaloid. A 2026 meta-analysis of 23 trials found an average weight difference of 0.88 kg against control. No source cited here reports a trial of berberine with a GLP-1 medicine.

Does berberine help you lose weight?

Slightly, on average, in the pooled trials. Meta-analyses found small reductions in weight and BMI, the results differ between reviews, and many of the trials were at high risk of bias. A combination capsule cannot show that berberine alone caused the change.

Is berberine safe?

EFSA’s panel could not establish a safe intake for any of the 13 plant preparations it assessed. It cited laboratory evidence of genotoxicity, interaction through CYP450 enzymes, carcinogenic activity of goldenseal preparations in rodents with a human-risk conclusion, and liver injury linked to greater celandine. For the other preparations, the toxicity profile beyond berberine is largely a data gap, not a certificate of safety. The UK Committee on Toxicity said the gaps reduced confidence in EFSA’s conclusions. It did not say berberine is safe. ANSES says to avoid these supplements in pregnancy, while breastfeeding, in children and adolescents, and if you have diabetes or a liver or heart disorder. Short-term complaints in the slides are mainly gut symptoms.

Is berberine being banned?

The sources cited here report no ban. The EFSA text is a public consultation, not a final opinion and not a law in the UK, US, UAE or South Africa. Risk managers in the Commission and member states decide whether to restrict the plant preparations. In July 2026 a regulatory expert quoted by NutraIngredients expected a final opinion in early 2027. That is a reported expectation, not an adopted rule.

How do I check a berberine product in Dubai?

Scan the barcode in Dubai Municipality’s Montaji app to see whether the supplement is registered in Dubai. Registration is not evidence that it works, and it is not evidence that it is safe with medicines.